EDITORIAL BRIEFING
Recent paths through the Jev ecosystem
Six records from the latest research batches show how bounded decisions are being used in coding tools, interactive interfaces and independent model work.
Coding agents
Recent projects put a judgment step between a coding agent and its next action: one narrows code-search leads, while another reviews proposed actions before the host decides what to do.
-
A CLI that uses Jev to judge relevance across repository folders, files and declarations, returning file leads and verbatim source excerpts to a coding agent.
integration · Batch 14 -
A research-stage coding harness where an LLM proposes an action, Jev answers four bounded questions and code records evidence for the host to review.
repo · Batch 13
Interactive control
Two demos apply Jev to constrained choices in a voice-driven browser and a driving simulation. Their records distinguish the model's decisions from the actions and safeguards implemented in code.
-
A local Node and Playwright app streams partial speech transcripts, asks Jev for intent and target choices, then applies code-level action gates.
demo · Batch 14 -
A playable Three.js driving simulation where Jev selects sampled steering and speed paths from compact road, traffic and route state.
demo · Batch 14
Tools and independent models
A local workbench packages versioned judgment functions; a separate Python library explores typed decisions from open models. They offer different ways to work with the same decision shape.
For every addition, see the changelog or subscribe to the Atom / JSON feeds. Individual records keep project destinations and original sources separate.